Deflation Prevention and Cure Willem Buiter Chief Economist 19 May 2003

1. DeflationPrevention and CureWillem BuiterChief Economist 19 May 2003

2. Introduction • It’s baaaaack! • Definition: sustained decline in cost of living; persistent negative inflation of price index for current goods and services. Practical examples: RPI, RPIX, RPIY, CPI, HICP, GDP deflator, private consumption deflator in the GDP index. • Asset price deflation; important, but not ‘deflation’ as used in this lecture.

3. Why worry about deflation? • (Alleged) historical association of deflation with financial crises, recession, stagnation and depression. Debt deflation and the Great Depression. • (Short) nominal interest rates at zero (‘zero lower bound problem’); liquidity trap – monetary policy cannot stimulate demand. • But: Bailey & Friedman’s optimal monetary rule: zero nominal interest rate. If real interest rate positive, deflation is optimal.

4. Japan: BoJ discount rate at 50bps or less since 1995; now near zero; slump since 1992, GPD deflator has fallen in each of past 5 years; CPI and money wages have fallen in 4 out of past 5 years. Why the concern today? A littlehistory

5. Japan

6. European Union: ECB’s repo rate at 2.5%; HICP inflation for 2002 2.2%; real economy faltering; euro strongest since introduction 4½ years ago.

7. European Union

8. USA: Federal Funds rate at 1.25%; real economy ‘soggy’ (Snow). CPI inflation 1.9% for 2002, PCE (personal consumption expenditures deflator) inflation1.9%.

9. USA

10. UK: Bank of England’s repo rate at 3.75%; For 2002: HICP inflation at 1.6%; GDP deflator at 3.2%; RPIX at 2.2%; Dec. 2002 at 1.9%, January 2003 at 2.6%, March 2003 at 3.0% (RPI at 3.1%); Sterling weakening; Real economy sluggish.

11. UK

12. UK 1800-1914: inflation highly volatile; average inflation rate negative

13. Bank rate never below 2%. Per cent US dollar Inflation (1hs) Bank rate (1hs) £/$ exchange rate (rhs)

14. Good Deflation and Bad Deflation (Bordo and Redish [2003])? • Good deflations: positive supply shocks cause potential real output to grow faster than nominal aggregate demand. Characterised by rising employment and output growth, high profits and booming stock markets. E.g. mild deflation during 1870 - 1896 accompanied by positive growth in many countries. However, growth accelerated during period of inflation after 1896. • Bad deflations: negative shocks to aggregate demand cause nominal demand growth to fall below the growth rate of potential real output. Characterised by falling employment and output growth, low profits and falling stock markets (1919-1921, 1929-1933, Japan 1995 -).

15. Is deflation just inflation with the sign reversed? Are anti-deflationary policies just anti-inflationary policies with the sign reversed? • Symmetries: • Menu costs • Shoe-leather costs: anticipated inflation (deflation) raises (reduces) opportunity cost of holding money if nominal interest rate adjusts one-for-one with expected inflation. • Asymmetries: • Zero lower bound on risk-free nominal interest rates • Unanticipated inflation redistributes from (non-index linked) creditors to debtors. Unanticipated deflation redistributes from (non-index linked) debtors to creditors. No counterpart in inflation case to default, bankruptcy and painful financial restructuring often found in the deflation case (debt deflation). • Asymmetries in upward & downward nominal wage and price adjustment • Unfamiliarity (outside Japan) with deflation in living memory.

16. Why don’t we see negative nominal interest rates? • Opportunity cost of holding base money: i-iM ishort nominal interest on non-monetary security iMnominal interest rate on base money • Bearer securities vs. registered securities • Monetary base: • bank reserves are registered securities • Currency is a bearer security • A ’carry tax’ on currency: Gesell’s Proposal.

17. Do asymmetric downward nominal price and wage rigidities make deflation particularly costly? • No nominal rigidities (either up or down) in asset markets. • In UK, with persistence of low inflation during past decade, price cuts for goods becoming common: UK, 2002 year to August, CPI inflation at 1.9%; CPI goods inflation at -1.1%; CPI services inflation at 4.6% (King [2002]). • Wages: Nickell and Quintini [2003] using micro-data find evidence of statistically significant but small nominal rigidities (spike at zero in density function of money wage contract changes at zero). Note: no evidence for or against asymmetry. • In inflationary world, resistance to a nominal wage cut is a necessary part of resistance to real wage cut. ‘Fairness’ concerns relative real wages.

18. Can we disentangle monetary and fiscal policy? Should we care? • 4 potential monetary instruments • Conventional • Short nominal interest rate on non-monetary securities • Stock of base money • Nominal spot exchange rate Note: of 3 conventional monetary instruments, only one can be independently selected by monetary authority under unrestricted international financial capital mobility. • Unconventional • Nominal interest rate on base money

19. Monetary Policy: any change in one or more of the 4 monetary policy instruments that, at given prices and activity levels, leaves unchanged financial net worth of state, now or in future (state = consolidated general government & central bank). • Fiscal policy: any change in public spending or tax rules, whether or not they alter, at given prices and activity levels, financial net worth of state, now or in future.

20. How to avoid deflation or how to escape from deflation once you’re landed in it. Amounts to question: how can monetary and fiscal policy boost aggregate demand. • Pedantic precision: how can monetary and fiscal policy boost aggregate demand at given (current and future expected) prices & wages, given current and future nominal exchange rates, given current and future expected employment and output levels.

21. (1) DH = CH + IH + GH + X DH :Aggregate demand for domestic output CH : Private consumption demand for domestic output IH : Private investment demand for domestic output GH : Government spending on domestic output X : Export demand

22. Share of private consumption of domestic output in total private consumption depends on the relative price of imports and domestic goods • Same for private investment spending on domestic output and government spending on domestic output. • Total private consumption is the sum of consumption by ‘Keynesian’ consumers who spend their current disposable income and consumption by ‘permanent income’ consumers, who smooth consumption over the life-cycle.

23. Aggregate consumption, C, is sum of the consumption of the permanent income consumers and the Keynesian consumers. • Share σ of consumers is Keynesian. The wage bill is W and labour income taxes is T. • (2) • Consumption by permanent income consumers is product of marginal propensity to spend out of comprehensive wealth, μ, and comprehensive wealth, sum of financial wealth, A, and its share 1-σ , of human wealth, H. C=µ[A+(1- σ)H]+ σ(W-T)

24. Financial wealth, A, includes stock market wealth, present discounted value of future profits. • Human wealth, H, is present discounted value of current and future after-tax labour income. • Private investment also depends on present discounted value of future profits (‘Tobin’s q’)

25. Monetary Policy • Until further notice, assume interest rate on base money (currency) is zero. • (1) Cut in current short nominal interest rate (consider effect on demand at constant exchange rate). • Effect on investment through Tobin’s q. Other mechanisms: credit channel; bank lending channel. • Effect on consumption through • Substitution effect • Income effect • Valuation effect. Substitution and income effects work through μ. Valuation effect works through A (stock market) and through H. Significant effect on μ unlikely (could be negative!). Any positive effect likely to be small (μ is small).

26. (2) Credible announcements of future cuts in nominal interest rates (can be associated with (1)). • (3) Devaluation (consider effect on demand at given current and future nominal interest rates). Marshall-Lerner conditions; expansionary effect strengthened if country is net foreign currency creditor. • (4) Unconventional monetary policy when short nominal interest rates are zero.

27. (a) Generalised open market operations • Purchases of any government securities with positive nominal interest rate • Sterilised foreign exchange purchases • Purchases of other foreign currency assets • Purchases of private domestic securities or real assets (stocks, bonds, options, other derivatives, property). • Easing of eligibility requirements for securities acceptable as collateral in repo operations. Expand list of eligible couterparties. Problems: moral hazard, adverse selection, governance. • (b) Spitting in the wind: introducing an inflation target or raising an existing one when you’re flat on the floor.

28. (c) Negative nominal interest rates through a carry tax on currency. • Fiscal policies to stimulate aggregate demand • (1) Debt-financed (lump-sum) tax cuts • (A) With Keynesian consumers • (B) With permanent income consumers only • (a) Without Ricardian equivalence (debt neutrality) • (b) With Ricardian equivalence • (2) Temporary increase in public spending on domestic output, financed any which way.

29. (3) The fail-safe policy: Friedman’s helicopter drop of money. Equivalent to tax cut (or increase in government transfer payments) financed by printing base money. • (4) Feldstein’s proposal: use current indirect tax cuts and future indirect tax increases to tilt intertemporal terms of trade in favour of current consumption. Similar to temporary investment tax credit or temporary investment subsidy.

30. Tackling deflation: how much can the central bank do on its own? • Open market operations. May need help from Treasury to do the unconventional stuff. • Can Central Bank perform helicopter drop of money on its own? Central Bank is not fiscal agent. Requires Treasury assistance (tax cut or transfer payment financed through Treasury borrowing, either directly from Central Bank or indirectly, with Treasury borrowing in market and Central Bank buying Treasury debt in market.

31. Conclusion • There is no excuse for persistent unwanted deflation. Deflation can be avoided. If it has taken hold, deflation can be eradicated. • Ending unwanted deflation is technically trivial and politically attractive: tax cuts and/or spending increases.

32. Inflation is always and everywhere a monetary phenomenon (Friedman) • … and the price of bananas is always and everywhere a banana phenomenon (Mrs. Mildred Smith from Tooting Bec) • Money is always and everywhere a fiscal phenomenon (Sargent and Wallace). • Excessive, inflationary budget deficits are the reflection of unresolved social conflict (Joachim Marx).

33. Final Reflections • Unwanted, persistent deflation is always and everywhere evidence of unnecessary, avoidable macroeconomic mismanagement (including failure to coordinate monetary and fiscal policy). • It is not believable that governments have forgotten easy mechanics & attractive politics of boosting inflation/eliminating unwanted deflation.